Alpha-GPC

A choline compound

Also called choline alfoscerate, Alpha-GPC is a small molecule used in choline and membrane metabolism. It is chemically distinct from citicoline and other choline preparations.^1364,1365

Read the population

Alzheimer’s disease and focus in healthy students involve distinct populations and outcomes. The strongest interpretation stays with the people, formulation and outcome that were studied.^1374,1390

Patient evidence

74 of 100

The efficacy analysis included 74 of 100 randomized participants. ADAS-Cog improved by 2.34 points with treatment versus 0.97 with placebo; total MoCA did not significantly differ.¹³⁸³

A positive older trial

In 261 people with Alzheimer’s disease, the 180-day ADAS-Cog change was -3.20 with treatment versus +2.90 with placebo. The current abstract supports those means; the full safety and completion tables were inaccessible.¹³⁷⁴

Two important null results

A 100-person donepezil add-on trial missed its main cognitive outcomes. CONIVaD also found no cognitive advantage among participants with small-vessel disease. These studies belong alongside ASCOMALVA’s positive reports.^{1377,1385,1375}

The Russian studies

The three-year comparison had severe baseline imbalance and no blinding or randomization. The five-year publication is accessible only as an abstract. Both add regional evidence with substantial uncertainty.^1388,1389

Focus, training and mood

Focus tests

A 20-man study found a Stroop-task improvement, with null working-memory and other attention outcomes. A separate 20-person conference study found no significant cognitive, mood or performance benefit.^1392,1393

Exercise claims

Several small experiments reported selected strength or cycling benefits; others were null. The well-known seven-man report is a conference poster. An eight-man growth-hormone experiment measured a short-lived hormone response rather than muscle growth.^{1395,1396,1398,1399}

Depression research

A 49-person placebo-controlled trial in type 2 diabetes found no depression-score advantage after six months. CARTESIO’s 2026 pilot reported improvement in 15 older completers over eight weeks, without a control group.^1403,1404

Safety questions

The stroke association

A large Korean health-record study associated Alpha-GPC prescriptions with higher stroke rates. A later matched MCI analysis was null and contains a stroke-code labeling inconsistency. Neither study randomly assigned treatment.^{1409,1384,1386}

TMAO findings differ

A six-man crossover found TMAO rises after water-soluble choline sources. A separate four-week abstract reported no difference between supplements or over time. Both measured metabolites, leaving long-term cardiovascular outcomes unsettled.^1408,1411

Sleep and agitation

A Polish injectable-product label lists uncommon anxiety, agitation, insomnia and nausea. Its revision date is blank. Oral supplement studies provide limited information about the frequency of these effects.¹⁴⁰⁵

Safety decisions

EFSA assessed 203.7 mg/day for a specified novel-food use. China’s food specification permits up to 600 mg/day and excludes pregnancy and breastfeeding. Each decision applies to its stated product, route and intended use.^1372,1373

Common questions

01 What is Alpha-GPC?

Alpha-GPC is a small choline-containing molecule, also called choline alfoscerate. It supplies a component used in acetylcholine and cell-membrane metabolism. Its chemical identity is distinct from citicoline, choline bitartrate and phosphatidylcholine.^1364,1365

02 Is Alpha-GPC a peptide?

No. It is a glycerophosphocholine molecule with the formula C8H20NO6P and molecular weight 257.22 g/mol. Reptides covers several non-peptide compounds alongside peptides.¹³⁶⁴

03 Is it an approved medicine or a supplement?

That depends on the country and product. EMA records identify nationally authorised medicines in Italy and Poland, and Russian records list choline alfoscerate medicines. Canada has natural-health-product guidance. Food-ingredient decisions assess specified uses and formulations separately from treatment claims.^{1366,1367,1368}

04 Does FDA GRAS status mean it treats memory loss?

FDA raised no questions about a 2012 notifier conclusion for specified food uses. The letter explicitly says FDA did not make its own GRAS determination. That process assesses a proposed food use; a dementia treatment claim requires a different evidentiary and regulatory basis.^1369,1370

05 What happened to the newer FDA filing?

The 2024 filing proposed substantially higher food-use exposure. FDA questioned the exposure estimates and found the submitted safety support insufficient, then recommended withdrawal and a meeting. The notifier requested that evaluation stop. The filing did not produce another no-questions response.¹³⁷¹

06 What did EFSA conclude in 2026?

EFSA considered a specified soy-derived ingredient safe under its proposed conditions, at 203.7 mg of Alpha-GPC per day for people above age three, including pregnancy and breastfeeding. This was a novel-food safety opinion, with a defined ingredient specification and intake. It did not evaluate treatment effectiveness.¹³⁷²

07 Why do countries list different daily amounts?

They assess different products and uses. China’s 2024 new-food specification sets a maximum of 600 mg/day on a dry basis and excludes infants, pregnant people and breastfeeding people. Canada’s 2026 adult cognitive-function monograph lists 1.2 g/day for an oral natural health product. Neither number is a universal personal dose.^1373,1368

08 What is the strongest older Alzheimer’s trial?

A placebo-controlled trial enrolled 261 people with mild to moderate Alzheimer’s disease. At 180 days, ADAS-Cog improved by 3.20 points with Alpha-GPC and worsened by 2.90 points with placebo. The study used 400 mg three times daily. The currently accessible abstract supports those results; its full safety and completion tables are unavailable in the accessible source.¹³⁷⁴

09 Does adding it to donepezil help?

Results are mixed. ASCOMALVA reported benefits in people with Alzheimer’s disease and vascular brain changes. A separate Korean randomized trial of 100 people found no significant advantage on its main MMSE measure or ADAS-Cog at either 12 or 24 weeks. The results come from different participants and study designs.^{1375,1376,1377}

10 Why are there so many ASCOMALVA papers?

The program generated several reports from one cohort, including later behavior, depression and MRI analyses. They add detail about that study’s participants, but should be counted as one clinical program when judging independent replication.^{1375,1378,1379,1380}

11 Do reviews settle the question?

The reviews combine studies with very different designs. One cognition review includes multiple ASCOMALVA reports and highly inconsistent results across its standalone comparisons. A stroke review combines citicoline and Alpha-GPC, while several Alpha-GPC studies track recovery without an untreated control. Their large totals need careful unpacking.^1381,1382

12 Has it been shown to prevent Alzheimer’s disease?

Prevention remains unproven. The main randomized studies enrolled people who already had cognitive impairment or dementia. Observational studies of later dementia diagnoses can be affected by who receives treatment, how long they survive and how exposure is counted.^{1374,1383,1384}

13 What did the 100-person MCI trial show?

After 12 weeks at 600 mg/day, ADAS-Cog improved by 2.34 points with Alpha-GPC and 0.97 points with placebo, with p=0.048. The efficacy analysis included 74 of the 100 randomized participants. Total MoCA scores did not differ significantly. The small difference, missing participants and multiple measured outcomes make replication important.¹³⁸³

14 What did CONIVaD find?

CONIVaD randomized 62 people with small-vessel brain disease and cognitive impairment to nimodipine with Alpha-GPC or placebo. Among 48 completers, seven in each group lost at least two MoCA points over a year. Secondary outcomes were also null. Adherence to background nimodipine was poor.¹³⁸⁵

15 What about the large Korean dementia study?

A 2025 study of 508,107 people with mild cognitive impairment linked prescriptions to lower rates of later Alzheimer’s and vascular dementia diagnoses. It was an insurance-record study, with treatment groups that initially differed substantially. Its handling of time before treatment and other residual differences limit causal interpretation.¹³⁸⁴

16 Why are its stroke results hard to interpret?

The 2025 paper reverses the ICD-10 labels for cerebral infarction and intracerebral haemorrhage in its methods. The paper does not resolve whether this is a wording error or an analysis error. Its matched analyses found no significant stroke association, so the unmatched subtype estimates are a poor basis for a headline.^1384,1386

17 Is CARL a completed positive trial?

The available CARL paper describes a planned 60-person study in mild cognitive impairment with vascular injury. It specifies treatment, imaging and cognitive assessments, but presents no efficacy results. The registry and protocol belong to the same trial.¹³⁸⁷

18 What did the three-year Russian study find?

The 2020 report followed 62 relatives of people with Alzheimer’s disease. Cognitive scores improved in the 30 treated participants and declined in the 32 untreated participants. All 30 treated participants had baseline deficits; only eight of 32 untreated participants did, while the other 24 were age-normal. There was no randomization, blinding or placebo.¹³⁸⁸

19 Did the five-year Russian study confirm prevention?

The 2026 abstract describes 122 people, with 56 receiving repeated courses and 66 untreated. It reports better cognitive trajectories in the treated group. The full paper remains inaccessible, so baseline balance, attrition and possible overlap with earlier cohorts cannot be checked; the abstract alone does not establish a preventive effect.¹³⁸⁹

20 Are Russian and Japanese studies included here?

Yes. The audit includes Russian patient studies and medicine records, a Japanese student esports trial, and Japanese animal conference reports. Each entry keeps its language, design and reading depth visible so a small unblinded study carries the appropriate weight.^{1388,1390,1391}

21 Does Alpha-GPC improve focus in healthy people?

The evidence is small and inconsistent. A 20-man crossover study found better performance on a Stroop task after acute doses, while working-memory and other attention measures were unchanged. Another 20-person conference study found no significant cognitive or mood benefit. A reliable everyday concentration benefit has yet to be demonstrated.^1392,1393

22 Can it make me more motivated?

A two-week, single-blind study of 39 healthy adults reported a positive nighttime motivation measure. Its overall motivation result was a trend, and anxiety measures were null. This is an early signal from a small study with several outcomes.¹³⁹⁴

23 Does it improve gaming performance?

A two-week Japanese study analyzed 20 students assigned Alpha-GPC or placebo. The higher-dose group had a lower salivary stress marker after gaming. Subjective stress rose only in the placebo group, producing a statistical interaction between group and time. Heart rate and overall working-memory performance showed no clear improvement. Competitive gaming performance was not established.¹³⁹⁰

24 Has it been shown to treat ADHD?

The cited Alpha-GPC trials do not establish ADHD treatment efficacy. Most healthy-adult studies measure short cognitive tasks in volunteers. Those studies cannot answer whether someone with diagnosed ADHD has lasting improvements in symptoms or daily functioning.^1392,1394

25 Does it improve strength or workout performance?

Some very small studies found improvements on selected strength or cycling measures. Others were null, including a 48-man study that found no strength advantage over placebo. The studies differ in dose, exercise task and timing, and usually follow participants for hours or days.^{1395,1396,1397}

26 What is the famous seven-person exercise study?

It is a conference poster describing seven men in a crossover experiment. A 600 mg dose before exercise increased peak force in the reported comparison. Several other measures, including peak power, rate of force development and metabolic rate, were null. The tiny sample and brief format limit the finding.¹³⁹⁸

27 Does the growth-hormone result mean more muscle?

An eight-man experiment found a short-term growth-hormone rise after 1,000 mg, with later changes in free fatty acids and ketones. It did not measure muscle growth or long-term fat loss. Training adaptations require studies that measure those outcomes over time.¹³⁹⁹

28 Do multi-ingredient pre-workout studies count?

They can test the complete blend. A 30-cyclist study combined Alpha-GPC with BCAAs and citrulline; some performance measures improved, while 20 km completion time and average power did not. It cannot identify which ingredient produced a difference.¹⁴⁰⁰

29 What about taking it with caffeine?

A 26-person crossover study tested a blend containing 100 mg Alpha-GPC, 300 mg caffeine and numerous other ingredients. Some cognitive measures improved, while motivation and anxiety were unchanged. The trial tested the complete blend, so it cannot tell us what Alpha-GPC adds to caffeine alone.^1401,1402

30 Could it help depression?

A placebo-controlled study of 49 adults with type 2 diabetes found no significant depression-score advantage after six months. The 2026 CARTESIO pilot reported improved scores in 15 older completers over eight weeks, but had no control group. These findings leave treatment benefit uncertain.^1403,1404

31 Could it affect anxiety, mood or sleep?

The Polish Gliatilin 1000 injectable label lists anxiety, agitation and insomnia as uncommon reactions, along with nausea. Its blank revision date leaves the document’s current status uncertain. Healthy-adult oral trials provide little dependable information on sleep effects. The 39-person motivation study found no clear anxiety benefit.^1405,1394

32 What amounts have researchers used?

The studies use widely different exposures: acute healthy-adult experiments include roughly 200 to 1,000 mg, the recent MCI trial used 600 mg/day, and several patient trials used 1,200 mg/day. Participants, formulations and endpoints differ. The studies describe experimental exposures and provide no recommended regimen.^{1393,1399,1383,1374}

33 How is it supposed to work?

Alpha-GPC contributes choline to acetylcholine metabolism and components used in membrane phospholipids. Animal and laboratory studies examine these pathways, including effects in damaged brain tissue. A plausible pathway helps explain what to study; clinical trials determine whether people benefit.^1365,1372

34 How quickly does it enter the blood?

In a 48-person formulation comparison after 1,200 mg, measured plasma choline peaked at about 3.5 to 3.9 hours. That timing describes choline in this study. It does not tell us when a person will notice a cognitive effect.¹⁴⁰⁶

35 Do we know its human half-life or brain exposure?

The clinical formulation studies largely measure plasma choline, including choline already present before dosing. They do not directly establish the half-life of intact Alpha-GPC in the human brain. Precise brain-exposure claims need a different measurement.^1406,1407

36 Does food or the formulation matter?

Meals change background plasma choline, which complicates pharmacokinetic comparisons. Tablets, soft capsules and different choline compounds have been studied under specific conditions. The available data do not establish one best meal schedule or prove that every retail formulation behaves identically.^{1407,1406,1408}

37 What side effects were seen in the recent trials?

The MCI trial recorded mostly nonserious events, including indigestion in both groups, with no serious adverse events or adverse-event withdrawals. The small CARTESIO pilot recorded a headache and transient high blood pressure. These modest studies cannot give reliable rates for rare harms.^1383,1404

38 Does Alpha-GPC increase stroke risk?

A Korean insurance study of more than 12 million older adults associated prescriptions with higher stroke rates. After matching and adjustment, the reported hazard ratio was 1.43. Treatment was not randomly assigned, and illness prompting a prescription may also predict stroke. The study reports an association and cannot determine whether the prescriptions caused the higher rate.¹⁴⁰⁹

39 How can the two Korean stroke analyses disagree?

They studied different populations, periods and prescribing patterns. The large 2021 study found an increased association, while the matched 2025 MCI analysis was null. Both use observational health records, and the later paper also has an unresolved stroke-code labeling problem. A randomized safety trial would answer a different, stronger question.^1409,1384

40 What is the TMAO concern?

Gut microbes can turn some choline into trimethylamine, which the liver converts to TMAO. A six-man crossover study found TMAO rises after water-soluble choline sources, including Alpha-GPC. Mouse experiments also connect Alpha-GPC exposure to TMAO and atherosclerosis. The relevance of these findings to long-term clinical risk remains unsettled.^1408,1410

41 Did a four-week study find the same TMAO increase?

Its publisher-deposited abstract reports no significant TMAO differences between the tested supplements or across time. The abstract omits the participant count, exact doses and safety details, and the full paper was inaccessible. Short metabolite studies can disagree without settling cardiovascular risk.¹⁴¹¹

42 Can taking it every day be considered safe?

Daily use has been studied, but the evidence varies by dose, population and duration. Many controlled trials lasted weeks or months; longer observational reports carry substantial uncertainty. Rare or delayed harms are difficult to measure in the available trials, especially for healthy people taking it indefinitely.^{1372,1409,1389}

43 Has it been linked to kidney cancer?

A 2026 Korean matched cohort compared 81,970 users with 409,801 nonusers and found no significant kidney-cancer association: adjusted hazard ratio 0.95, with a 95% confidence interval of 0.84 to 1.08. The currently accessible abstract supports that specific outcome; it does not assess every cancer or safety concern.¹⁴¹²

44 What do we know about pregnancy and breastfeeding?

The documents assess different uses. China’s food specification excludes pregnancy and breastfeeding; Canada’s adult monograph advises a health-professional consultation. EFSA includes these groups only within a specified 203.7 mg/day food use. The Polish injectable label restricts pregnancy use to absolute necessity and advises against breastfeeding use. Evidence for higher-dose self-treatment remains insufficient.^{1373,1368,1372,1405}

45 Could it interact with medicines?

The Polish injectable label says interaction studies were not conducted. Some patient trials deliberately combined Alpha-GPC with donepezil or nimodipine under supervision. A medication review is appropriate before adding it to treatment that affects cognition, blood pressure or cholinergic signaling; the trials do not cover every combination.^{1377,1385,1405}

46 What does a 50% or 85% powder label mean?

It describes the proportion of the preparation that is Alpha-GPC. For example, 600 mg of a preparation genuinely containing 50% Alpha-GPC supplies 300 mg of Alpha-GPC. Whether the label already expresses active content must be checked before comparing a product with a study.^1372,1390

47 Is purity just about the percentage on the label?

No. Quality assessment also examines related compounds, stereochemistry, water, residual materials and other impurities. Chinese analytical papers describe methods for checking these issues; a 2026 series found frequent specification failures for some related substances. These were ingredient samples, so the findings cannot rank retail brands.^{1413,1414,1373}

48 Does soy-derived Alpha-GPC matter for allergies?

The EFSA-assessed ingredient was made from soy phospholipids, and its opinion addresses allergen labeling. Other production routes also exist. Someone with a soy allergy needs the specific product’s source and allergen information; the chemical name alone does not supply it.^1372,1373

49 Is it prohibited in sport?

Alpha-GPC is not named on the 2026 WADA Prohibited List or Monitoring Program. Athletes still need to check the complete product and administration method. The list separately restricts certain intravenous infusions and injections, and a supplement label cannot guarantee the absence of undeclared substances.^1415,1416

50 Was Alpha-GPC proven to cause a doping case?

A UKAD case mentioned an Alpha-GPC supplement among several alleged exposures after an ostarine finding. The tribunal did not establish the source. That decision should not be presented as proof that Alpha-GPC itself is prohibited or that the named supplement was confirmed contaminated.¹⁴¹⁷

51 Is Alpha-GPC better than citicoline?

There is no reliable general ranking. A 2025 review found three older injected-drug comparisons in patients, all with high risk of bias. A global clinical scale favored Alpha-GPC, while memory and word-fluency outcomes did not. Those trials cannot choose the better oral supplement for a healthy person.¹⁴¹⁸

52 What is the fairest overall verdict?

Alpha-GPC has patient trials, nationally authorised medicines and plausible biology. Clinical results are mixed, healthy-adult studies are small, and longer-term safety questions remain. The most useful judgment is specific to the outcome: a memory score in a patient trial, a short exercise test and a stroke association are different kinds of evidence.^{1374,1377,1392,1409}

53 What research would change that verdict?

Larger independent trials with prespecified outcomes, complete follow-up and clear adverse-event reporting would help most. For healthy users, a brief task score leaves everyday functioning and long-term safety unanswered. Existing trial registrations and protocols should be followed through to published results.^{1383,1392,1387}